The Effect of Task Repetition and Task Recycling on EFL Learners' Oral Performance

The recent literature on second language acquisition (SLA), particularly within classroom context, has seen a growing interest in the use of tasks. Task-based language teaching (TBLT), as Ellis (2009) puts it, has become popular with both researchers and teacher educators. It has drawn extensively on research into second language acquisition as reflected in seminal edited collections (e.g., Bygate, Skehan, & Swain, 2001; Crookes & Gass, 1993; Ellis, 2005; Garcia Mayo, 2007;  Willis & Willis, 1996) as well as recently published research articles in both EFL and ESL contexts (e.g., Ahmadian, 2012; Ahmadian & Tavakoli, 2011; Baleghizadeh & Nasrollahi Shahri, 2013; Tavakoli & Foster, 2011). In addition to SLA researchers, teacher educators have also presented a strong case for it (e.g., Estaire & Zanon, 1994; Nunan, 1989, 2004; Willis, 1996). Both SLA researchers and teacher educators emphasize the vital need to elicit samples of language use that are representative of how learners perform when they are not attending to accuracy (Ellis, 2003). Tasks have proved to be the right means providing these samples of meaning-focused language use. Tasks provide input to learners and opportunities for meaningful language use, both of which generally considered valuable in promoting language acquisition (Swain, 1995). Hence, tasks hold a central place in current SLA research and also in foreign/second language pedagogy.

    Tasks have played a key role in both the early descriptive research and the later more theoretically based SLA research and have been in the spotlight of research in their own right (Ellis, 2003). In fact, TBLT has gained such importance that van Lier (1994) believes it is an area where SLA and pedagogy are interdependent pursuits. However, it should be acknowledged from the outset that in neither research nor language pedagogy is there complete agreement as to what constitutes a task, nor is there consistency in terms employed to describe the different devices for eliciting learner language. As Samuda and Bygate (2008) note, forming such a definition is anything but easy. These definitions have encompassed a number of different dimensions and aspects to the concept of task, the scope of a task, the perspective from which a task is viewed, the authenticity of a task, the linguistic skills required to perform a task, psychological processes involved in task performance, and the outcome of a task (Ellis, 2003). Over the last few decades, different definitions have been proposed (Ellis, 2003; Nunan, 1989; Prabhu, 1987; Skehan, 1996; Samuda & Bygate, 2008) with some restricting the use of the term to activities where meaning is primary while the others have emphasized that learners will need to pay attention to both meaning and form in both tasks and exercises (Widdowson, 1998).

    As for providing learners with purposeful learning experiences, tasks differ from other approaches in several ways in. As Willis (1996) observes, many mistakenly view tasks as an umbrella term for various activities including grammar exercises, practice activities, and role plays. Nevertheless, tasks are different from traditional classroom approaches. Tasks mainly focus on meaning and form and inspire learners to approach language learning as a purposeful experience.

 

LITERATURE REVIEW

Since its inception, TBLT has been an exciting approach to second/foreign language pedagogy (Lynch & Maclean, 2000) and has become popular with both researchers and practitioners (Swan, 2005). Many publishers and materials writers have sought solace in the catchy motto of TBLT with the presumption that it would boost the process of second language  (L2) learning and it would ultimately lead to better outcomes in terms of what learners have actually gained through performing tasks.

    In contrast with the dazzling glamour of tasks, however, there have been certain criticisms against the claims TBLT makes about how tasks promote learning. In reality, it has been the center of many criticisms, a few of which have been well-grounded while the majority, as Ellis (2009) rightfully observes, stem from a fundamental misunderstanding of what a task is and of the theoretical bases that inform TBLT. These criticisms also reflect a failure to acknowledge that multiple versions of TBLT exist. In particular, it is argued that TBLT need not be seen as an alternative to more traditional, form-focused approaches but can be used alongside them.

    There exists a trade-off between a focus on meaning as advocated by TBLT and a focus on form as said to be subordinated in tasks viewed as a downside of TBLT. One of the best ways to counteract the negative effects of such a trade-off is to encourage learners to have a shift of attention from meaning to form (Willis & Willis, 2007). As Bygate (1996) observes:

 

This shift, from a preoccupation with finding the expressions to a greater capacity for monitoring formulation, may be precisely what teachers might wish to encourage since it may enable learners to pay more attention to the task of matching language concepts, and possibly to improving their knowledge and organization of the language (p.144).

 

    The post-task phase affords a number of options to counteract the danger that students will develop fluency at the cost of accuracy, one of which is to repeat the task a second time, namely task repetition, and the other is assigning a similar task having the same communicative goals, namely task recycling with results differing from one study to another (see Bygate, 1996; Bygate, 2001; Bygate & Samuda, 2005; Gass et al., 1999; Gass & Varonis, 1985; Lynch & Maclean, 2000; 2001). Task repetition and task recycling were presumed to positively affect fluency, accuracy, and level of complexity of students but the results gained from the studies proved not to be in congruity with the theoretical merits of task repetition and recycling on students' second performance. One crucial factor to bear in mind is to do with the fact that no previous study on any possible effects of task repetition/recycling has included both types as long as the number of task participants is concerned. Hence, the literature is not clear about any possible difference in students' performance when doing a task individually as opposed to pair work tasks. Besides, pair work tasks were preferred over group work tasks since, as Foster (1998) found out, there was more negotiation in the pairs than in the groups irrespective of tasks. Furthermore, it is easier to analyze the participants' speech as the voice recorded would be more intelligible compared with the case of group work tasks when the recorded speech would be garbled.         

    Given the few number of short-scoped studies on task repetition and task recycling, one should be cautious about the results of the studies on the effect of task repetition and recycling on the learners' performance. For one thing, the results found in these studies lack generalizability since they mostly focus on one single participant (Bygate, 1996), or case studies of 2 or 3 participants (Bygate, 2001; Bygate & Samuda, 2005; Gass et al., 1999; Lynch & Maclean, 2000, 2001). Furthermore, the research intervention has been brief as it does not reflect conventional extended pedagogic involvement, as would be regarded as typical by language teachers. In other words, unlike the case of poster carousel in Lynch and Maclean studies (2000, 2001) where the participants were medical students and the task of explaining about their medical posters was similar to a real life task in medical conferences, the treatment of the other studies has not been a part of ordinary classroom activity. Thus, one of the novelties of the present study pertains to the fact that the tasks have been a part of the usual classroom activities, and not a separate activity, which greatly reduces the problem of artificiality of the tasks in this study. In addition, there are few studies, if any, that have scrutinized the different ways in which task repetition and task recycling affect students' performance; nor does exist any serious attempt which looks into any possible discrepancy or congruity of student's oral performances when doing task repetition or task recycling individually or in pairs.

 

PURPOSE OF THE STUDY

The present study aims to scrutinize the effects of task repetition and task recycling on learners' fluency, accuracy, and level of complexity. The comparison is made on two levels, differences between individual and pair work performance of learners within task repetition and task recycling alongside with the inter-group differences of both the task repetition and task recycling groups to find out which one, in practice, outweighs the other. To this end, the present study seeks to answer the following research questions:

 

1. Does task repetition, carried out individually or collaboratively, have a significant effect on learners' oral performance in terms of fluency, accuracy and complexity?
2. Does task recycling, carried out individually or collaboratively, have a significant effect on learners' oral performance in terms of fluency, accuracy and complexity?
3. Is there any significant difference between learners' performance in task repetition and task recycling in terms of fluency, accuracy and complexity?

 

METHOD

The present study employed a quasi-experimental research design since it lacks randomization, yet employs other strategies to provide some control over extraneous variables. It is important to note that the number of participants in this study is not enough as recommended for quasi-experimental studies. Rather, the present study is a case study very similar to the previous studies on task repetition/recycling given the time constraints of the research and the fact that data collection of this type involves transcribing the oral production of participants which requires a team of researchers to be able to both quantitatively and qualitatively analyze the participants' production. In their oral productions, the dependent variables of the study, namely fluency, accuracy, and complexity were measured and then statistically analyzed using the Wilcoxon signed rank test.

 

Participants 

Initially, the participants for the present study were 30 Iranian learners of English at intermediate level, whose age ranged from 22 to 34. They were both male and female studying Master Class FCE (2008) at Iran-Australia School of Foreign Languages in Tehran. The participants, members of two intact classes each consisting of 15 students, were selected as two experimental groups with no control group. The students in the first class were assigned to do task repetition and the students in the second class were asked to do task recycling. In each group, four students whose oral productions formed the basis of the study were randomly selected. It should be noted that although the participants had already been assigned to their classes based on the institute's initial placement test and final exams at the end of each term, they were given the speaking test of IELTS in order to have a clearer picture of their speaking command prior to the actual treatment.

 

Instrumentation

The study made use of eight different opinion gap tasks. They fell under the category of opinion gap tasks since they involved learners in going beyond the information given by supplying their own ideas, students had all the information at their disposal, and the information exchange was optional (Ellis, 2003). The reason for choosing opinion gap tasks and not information gap tasks lie in the fact that the conversational task, as Nakahama, Tyler and van Lier (2001, cited in Ellis, 2003) concluded, offers a larger range of opportunities for language use.

    There were four pictorial tasks and four tasks with verbal instructions (verbal input medium). The reason why two different input media were chosen was to eliminate any extraneous factors having to do to with task complexity factors. Moreover, since the participants had already passed the course book Countdown to First Certificate (OUP, 2008) during their Pre-FCE Course, they were quite familiar with such tasks. Pictorial opinion gap tasks were chosen because textual input that is supported by visual information in some form is generally easier to process than information with no such support (Ellis, 2003). In the case of tasks with verbal rubrics, prompts were provided as to reduce the level of complexity of tasks. Furthermore, more time was allotted to verbal tasks since, as Nunan (1989) notes, the texts supported by photographs, drawings, tables, and graphs are easier to understand.

    The first two pictorial opinion gap tasks involved four common problems: poverty, air pollution, smoking, and garbage, for which the participants were asked to discuss the negative tentacles of each problem on people's lives and the society. The third pictorial opinion task involved two pictures depicting borrowed cultures in oriental and occidental countries which required the participants to discuss the advantages and/or disadvantages of crossing cultures. The fourth pictorial opinion gap task illustrated two group activities versus two individual activities and participants were asked to discuss the upsides and/or downsides of group activities and individual ones.

    The other four tasks had verbal instructions with prompts to lessen the level of complexity. In the first two ones the participants were asked to discuss and elaborate on the professional qualifications and characteristics needed to hold certain jobs. The other two tasks, however, required them to give suggestions for improvements needed in different areas in the city and the university campus respectively. It is worthy of mention that the instructions of all eight tasks required students to give reasons to support their viewpoints (for a sample of tasks, see Appendix A). Four of the tasks required the participants to work in pairs, whereas in the other four tasks, they worked individually.

 

Data Collection Procedure

In the first stage, two intact classes of equal number of participants were assigned to two experimental groups labeled as task repetition and task recycling, respectively. Then a pretest, a version of IELTS speaking test, was administered by the second researcher and his colleague to the participants on both groups to have a clear picture of their level of English language proficiency. To this end, the second researcher and his colleague gave them a speaking test on two separate occasions to remove the threat imposed by experimenter effects since the second researcher was their teacher and he intended to have a more reliable pretest. He and his colleague were both experienced in teaching IELTS with eight and ten years of teaching experience, respectively. After the pretests were administered, the participants' scores were compared and the inter-rater reliability was calculated and the results indicated a high reliability index of 0.92.

    The whole data collection process took four consecutive sessions. The reason for such intervals was to counter any possible effects of history and maturation. The participants in both groups were required to perform a given task at the beginning of the class as part of their regular class work assigned by their course book and their performance was subsequently recorded. Their second performance of the same task in the task repetition group and of a similar task in the task recycling group was again recorded toward the end of the class. In each group, four out of 15 participants were randomly chosen and their performances formed the basis of analysis for this study. However, none of them were aware of the fact that their performance is a part of the study, lest this would have had an impact on the dependent variables of the study, namely fluency, accuracy, and the level of complexity of the their performance. The basic procedure of the study is demonstrated in Figure 1:

 

	Group A (Task repetition)	Group B (Task recycling)
Session one	The first pictorial pair work task on social problems with its repetition	The first two pictorial pair work tasks on social problems
Session two	The fourth pictorial task on crossing cultures with its repetition done individually	The third and the fourth pictorial tasks on crossing cultures and activities done individually
Session three	The first verbal instruction task on city improvements done in a pair	The first and second verbal instruction tasks on city and university improvements done in a pair
Session four	The third verbal instruction task on professional qualities and its repetition done individually	The third and the fourth verbal instruction tasks on professional qualities done individually

Figure 1: Data collection and task types in the study

 

    The recording process utilizing a built-in microphone and closed-circuit camera installed in all classes was a routine in the language school with which the participants were quite familiar. To enhance sound quality, a professional recorder was also used which was stuck underneath the chair placed in the center of the class.

    As long as the time allotment of tasks was concerned, six minutes was allocated to pictorial tasks and eight for verbal instruction tasks. The same time was allotted for the second performance of the two groups. No corrective feedback, implicit or explicit, was provided by the teacher, especially in the first performance in order not to put the second performance of the participants at an advantage. Between the first and second performance of the tasks, class would continue covering its ordinary syllabus. Toward the end of class, students were asked to repeat the tasks in the task repetition group or perform a task of similar content, communicative goals and rubrics in the task recycling group. Although there was fear that the second performance of the task may have created artificiality, it must be borne in mind that all the tasks incorporated in this study were a part of classroom activities interwoven into the syllabus of the class. Besides, as Bygate and Samuda (2005) argue, teachers should clearly consider building repetition into their classes since it appears to be a fundamental characteristic of ordinary discourse. If they asked the reason, the teacher would justify the second performance under the pretext that the first performance had not been satisfactory. No elaboration, however, was furnished on their syntactic, lexical, or pragmatic errors or mistakes, whatsoever. On no occasion had the participants had any practice or opportunity to prepare for the task, since the instruction for the second performance was totally unexpected.  

    At the end of each treatment session, the participants' recorded performance was transcribed by the second researcher. Then, the frequency count for the dependent variables, namely fluency, accuracy, and complexity (CAF) was calculated and compared in each group individually and as well between the two experimental groups (repetition and recycling groups) to see if there is any difference between task repetition and task recycling in the way they affect CAF. Besides, the same distinctions were made between CAF measures calculated for individual tasks as opposed to pair works. The operationalized measures of CAF will be discussed in the data analysis.

    Before the main study was conducted, a pilot study had been run to detect any likely unprecedented pitfalls in the implementation of the design, time allotment, and the interpretation of the data under scrutiny. One significant decision to be made was to do with the task type and check whether or not the particular ones used in previous studies would be practical. Several drawbacks were spotted when piloting the video narration. The first and foremost was that the cartoon narration used by Bygate (1996, 2001) could not be used here as the participants in this study were adults and many of them would simply not buy it. Instead, one episode of National Geographic series and another from BBC documentary series were put to test in two different FCE classes different from those included the study. The task was to summarize and report what they watched. On two occasions, however, there seemed to be a recurring pattern. Students were more interested in watching the video and knowing exactly what was being played and commentated on the video. The problem was, in many instances, they would pause and ask the researcher the specific word or term used in the text. When they were not provided with the answer by the researcher, they would get distracted by either asking others or checking their dictionaries which thwarted the very goal of the task. 

    Poster carousel task used by Lynch and Maclean (2000, 2001), involving medical students, was modified and piloted as well. This type of task was changed to the one the researcher found to be of more interest to students. Students were asked to create a poster-like job profile of the jobs they held at the time and include the characteristics needed. Although the majority of students seemed to have liked the task, one major flaw was distinguished. While students were walking around the class asking and answering about the posters, they joked around and used their mother tongue which created a hodge-podge difficult to analyze. Students were asked on several occasions not to get off-task, but the problem still perpetuated though minimized. It is important to note that the task used in their studies had resemblance to what the participants did in reality since they were medical students preparing for a medical conference whereas a modified version of their poster carousel piloted here had an obvious problem of artificiality which apparently led to the problem discussed above.

    Therefore, two types of opinion gap tasks in which the participants were required to express their opinions and furnish reasons with different input media were chosen for the present study. The materials employed consisted of four pictorial and four verbal instruction tasks with prompts. Half of the tasks employed in this study were individual tasks and the other half were pair work tasks.

 

Data Analysis

The participants' oral performances on all eight tasks were transcribed. They were randomly minimized (20 percent) in order to check the reliability indices of accuracy, fluency, and complexity. For this purpose, two raters examined the samples. They were both experienced EFL teachers with 10 years of teaching experience. The data were examined and the result of inter-rater reliability for accuracy, fluency, and level of complexity of the participants' oral performance indicated 92.1%, 94%, and 90.3% agreement respectively. Given this high inter-rater reliability, the remaining data were examined by the researchers. The raters followed the rubrics below while judging the fluency, accuracy, and complexity of the participants’ oral performance: Fluency was calculated considering the speech rate with the number of words per minute (Richards & Schmidt, 2010); accuracy was calculated by measuring the percentage of error-free clauses, or errors (syntactic, lexical, and pragmatic) per 100 words (Skehan, 2003); and complexity was measured by computing the average T-unit length by counting the number of T-units and dividing them by the total number of words produced (Wolfe-Quintero, Inagaki, & Kim, 1998)

    To analyze the data, Wilcoxon signed ranks test was implemented. The reason for that lies in the fact that the data collected in the study were of non-parametric nature, since they were frequency counts of the dependent variables (CAF). Although in the case of fluency and accuracy there were equal intervals, no mark could be assigned to them in order to measure and analyze the data using one of the parametric tests like t-test or ANOVA. It is important to bear in mind that since the two groups in the pilot study as well as the main study were at the same level of proficiency and were homogenized prior to the actual treatment, the use of Wilcoxon signed ranks test was statistically justified.

For the ease of reference, the following abbreviations will be used in the rest of the article:

 

• F1 & F2 are 1^st & 2^nd fluency measures of the task repetition performance;
• FF1 & FF2 are 1^st & 2^nd fluency measures of the task recycling performance;
•  C1 and C2 are 1^st & 2^nd complexity measures of the task repetition performance;
• CC1 & CC2 are 1^st & 2^nd complexity measures of the task recycling performance;
• A1 & A2 are 1^st & 2^nd accuracy measures of the task repetition performance; and
• AA1 & AA2 are 1^st & 2^nd fluency measures of the task repetition performance.

 

RESULTS

Investigating Research Question 1

The first research question deals with the possible effects of task repetition on learners' individual/collaborative performance with a change in their complexity, accuracy, and fluency (CAF) from the first attempt to the second. For this purpose, the measures of CAF in task repetition group are separately analyzed for individual tasks and pair work tasks. For both individual and collaborative performance in task repetition, it was hypothesized that the second performance would be advantaged with more fluency (i.e. more number of words per minute); more accuracy (i.e. fewer number of lexical and syntactic errors), and more complexity (i.e. higher ratio of t-units divided by the total number of words produced).

    As Table 1 reveals, the individual performance of task repetition does make a change in terms of CAF. The change from the first performance to the second in task repetition group is for better improving the CAF of participants with .031, .033, and .034 all smaller than .05 indicating a significant change. Therefore, the first research question for task repetition in individual performance is answered positively. In other words, it can be concluded that task repetition leads to a significant increase in the learners' oral individual performance in terms of CAF.

  

Table 1:  Wilcoxon signed ranks test for CAF in individual task repetition

F2-F1                A2-A1                 C2-C1

Z                   -1.826                 -1.841                  -1.857

Sig                 .034                    .033                      .031

 

    According to Table 2, the pair-work performance of task repetition as well led to an increase in learners' oral performance in terms of CAF. The significant change as indicated by .033, .034, and .032 for complexity, accuracy, and fluency suggests that the first research question for task repetition in pair work performance is answered positively as well. In other words, it can be concluded that task repetition results in a significant increase in the learners' oral pair work performance in terms of CAF.

    Comparatively speaking, the fluency of pair work performance of task repetition appears to be slightly higher than that of individual tasks. In the case of complexity, and accuracy, individual task repetitions seems to result in a minimally better performance than in pair work task repetition.

 

Table 2: Wilcoxon signed ranks test for CAF in pair work task repetition       

F2-F1                  A2-A1                 C2-C1

Z                        -1.841                   -1.826                 -1.841

Sig                      .032                      .034                     .033

 

    Table 3 demonstrates the statistical performance of the participants in the task repetition group, namely the aggregate of the measures of dependent variables in their performances in individual task repetition plus pair work. On the whole, as revealed Table 3, the first research question is completely answered in the positive given the fact that the second performances of task repetition participants were more fluent, more accurate, and more complex.

 

 

 

Table 3: Wilcoxon signed ranks test for CAF in task repetition

F2-F1                  A2-A1                 C2-C1

Z                        -1.826                   -1.841                 -1.890

Sig                      .034                      .032                     .029

 

Investigating Research Question 2

According to the second research question, it was hypothesized that task recycling affects learners' oral performance in terms of CAF when done both individually collaboratively. Very similar to the approach when investigating the first research question, the participants' performances were scrutinized on individual as well as pair work levels.

     Unlike the case of the task repetition group, as Table 4 depicts, task recycling did not lead to an overall increase in the oral performance of the learners when done individually as no significant change was observed in the accuracy and complexity of participants' second performances in task recycling. The only area affected by a significant improvement is fluency with the significance of .034. The results obtained from Table 4 add a few wrinkles to the second research question, as two measures of oral performance did not have any meaningful change and it was only fluency which actually improved. Research question 2 is therefore answered negatively regarding individual task recycling.  

 

Table 4: Wilcoxon signed ranks test for CAF in individual task recycling

FF2-FF1            AA2-AA1              CC2-CC1

Z                        -1.826                   -1.069                   -1.414

Sig                      .034                      .142                       .078

 

    When it comes to the effect of task recycling on learners' pair work performance, as illustrated in Table 5, it is clear that there are no significant changes contrary to what was initially hypothesized, that is a second performance of a similar task of the same type where communicative goals remain the same, namely task recycling would improve the participants' performance. In this case, there is no one single significant change in any of the dependent variables unlike the individual performance where fluency witnessed a meaningful rise. As for the case of fluency, the obtained figure is .054, which is very close to the margin of significance but not close enough to be interpreted as a viable one.

 

Table 5: Wilcoxon signed ranks test for CAF in pair work task recycling

FF2-FF1            AA2-AA1              CC2-CC1

Z                        -1.604                   -1.000                   -1.000

Sig                      .054                      .158                       .158

 

      A brief look at the overall within group comparison of performances in the task recycling group (see Table 6) reveals the fact that with the exception of a meaningful change in the fluency of the second performance of the participants in this group, task recycling, on the whole, does not affect the oral performance of the learners. Thus, the second research question is answered negatively here, since two measures of oral performance have not significantly changed.

 

Table 6: Wilcoxon signed ranks test for CAF in task recycling

FF2-FF1            AA2-AA1              CC2-CC1

Z                        -1.841                   -.816                     -1.342

Sig                      .033                      .210                        .089

 

Investigating Research Question 3

The most novel part of the present study is related to the third research question, which deals with the performance difference(s) in the learners' oral performance in terms of CAF affected by task repetition and task recycling. The rationale supporting such a claim pertains to the fact that, to the best of the researchers’ knowledge, to date there has not been a single study delving into the possible effects of task repetition and recycling in the literature. The analyses of the previous two research questions made it clear that task repetition is superior in affecting learner's oral performance compared with task recycling. In what follows, however, the statistical difference between the task repetition group and the task recycling group will be provided. To begin with, the differences between the effects of task repetition and task recycling on CAF both individually and collaboratively will be furnished. Then, they are followed by the overall differences of both groups.

    It was initially hypothesized that the task repetition group would prove to be superior and its improvement meaningful compared with the task recycling group in both individual and pair work performance as well as overall performances. According to Table 7, in all three performance measures, namely complexity, accuracy, and fluency, the second individual performances of the task repetition group appears to have a significant improvement in comparison with those of the recycling group with .034, .033, and .034, respectively. It can, hence, be concluded that there sure is a difference between oral performance of learners in terms of CAF when doing task repetition and task recycling with task repetition outweighing task recycling in terms of CAF.

 

Table 7: Wilcoxon signed ranks test for between group individual differences

DFF-DF            DAA-DA                DCC-DC

Z                        -1.826                   -1.841                     -1.826

Sig                      .034                      .033                         .034

 

    Results of Table 8 for between-group differences in participants' pair work performances are in congruity with those of individual between-group differences. Here, as well, a significant difference can be observed between the performances of task repetition and task recycling group with an advantage for the former for all three dependent variables with .033, .033 and .032 for CAF all smaller than .05.

 

Table 8: Wilcoxon signed ranks test for between group pair work differences

DFF-DF            DAA-DA                DCC-DC

Z                        -1.841                   -1.841                     -1.841

Sig                      .032                      .033                         .033

 

    As Table 9 clearly demonstrates, there is a clear distinction between the oral performance of learners doing task repetition and task recycling with, again, a significant difference in favor of the task repetition group. The conclusion, therefore, can be drawn that research question 3 is answered positively given the fact that there is a significant difference in all three between group variables with .031, .034, and .034 for CAF respectively.

   

Table 9: Wilcoxon signed ranks test for between group differences

DFF-DF            DAA-DA                DCC-DC

Z                        -1.826                   -1.826                     -1.857

Sig                      .034                      .034                         .031

 

DISCUSSION

One crucial factor is that there was almost no study in this area comparing task repetition and task recycling, nor was there any study conducted on within group differences of individual or collaborative performance of tasks. The rationale behind the previous studies drew upon the limited attentional resources noted by Skehan (1996). In other words, to counteract the drawbacks of too much focus on meaning, achieving fluency at the expense of accuracy (Willis, 1996; Willis & Willis, 2007), neglecting a focus on form, many experts advocate a focus on form once the task is completed (e.g., Bygate 1996, 2001; Ellis 2003). Willis and Willis (2007)  urge teachers to make sure that this focus on form does not detract from a focus on meaning and believe the simplest way to do this is to withhold focus on form until after a task has been completed.

    Considering the first research question on task repetition and its possible effects on learners' oral performance in terms of CAF, it was concluded that it did have a positive effect on all three of them. One reason to explain this has to do with task and content familiarity learners gain when doing a task for the first time. During task repetition, the cognitive load of the tasks is substantially reduced as learners are familiar with not only the content, but also the task instructions. They can attend to accuracy, complexity and produce more fluent utterances. Concerning accuracy and complexity, they can retrospectively, draw upon the lexical and syntactic aspects of the language produced, both by themselves and the partners and/or others in the classroom.

    The findings of the first research question on fluency and accuracy see eye-to-eye with Bygate (1996), who also states that repetition leads to changes in learners' use of the language system, to increased fluency (as reported for both individual and collaborative tasks), and perhaps to increased awareness. Findings on complexity, accuracy, and fluency agree with those of Gass et al. (1999) maintaining when the focus on meaning has been minimized through task repetition, because of familiarity with the content, learners are freed from their own internal system, and can attend to form. The findings on task repetition effects do agree with Bygate (2001) on improvement in fluency and complexity measures. However, in the case of accuracy they do not, since the present study showed an improvement in all three areas while Bygate did not report any improvement in the participant's accuracy. As for Lynch and MacLean's (2000) study, the findings related to the first research question also agree with theirs in terms of accuracy, whereas there was no instance of phonological gains as they reported, since “mispronunciation” was not measured in the present study. They also had another case study (Lynch & Maclean, 2001) with three partners. In both studies, they concluded that different learners with different levels of language proficiency benefited in different ways (see Alicia and Daniela in the first study and Susanna, Olga, and Carla in the second study each having different gains). As for Bygate and Samuda's (2005) cartoon narration study (a follow-up of Bygate’s 1996, 2001) where they looked at the effects of task repetition on framing, the findings agree with their reporting on lexico-grammatical gains, and not with their framing measure.

    Considering the second research question concerning the possible effects of task recycling on the learner's oral performance, it was answered negatively except for a significant gain in fluency of individual task recycling performance as well as a meaningful improvement in fluency in the overall task recycling performance. There are certain crucial points to be discussed. The first and foremost notion is to do with the reasons justifying a gain only for the fluency measure and not the other two. One possible explanation for this can be drawn on Skehan's (1996) three aspects of task performance. He maintains meaning is reflected in fluency where form is manifested in either accuracy (if control is prioritized) or complexity (if opportunities for restructuring arise out of students' willingness to take risk). According to Skehan (1996), task demands push learners in certain ways, prioritizing one or another aspect of language; however, the key question here is when learners are confronted with demanding tasks. One possibility is that the learners' basic choice is between meaning and form. In the present study, the participants seem to have favored meaning over form given the fact that in the second performance they were already familiarized with the communicative goals and this made them inclined to have higher fluency compared with their first attempt. Skehan (1998) further maintains that learners not only have to choose between meaning and form, but also between accuracy and complexity. He further suggests that there is a trade-off between accuracy and complexity. Complex tasks may lead learners to adopt either a safety first approach by electing to use language for which they have already developed automatic processing in which accuracy increases at the expense of lower complexity or an accuracy last approach where they attempt to utilize language requiring controlled processing but are unable to pay sufficient attention to it in which complexity rises at the cost of less accurate performance. In this respect, however, the findings on task recycling performance did not adhere to what Skehan proposed. This claim does as well require a further analysis to check whether the findings of the present study are confirmed.

    One other interesting finding of the results concerning task recycling is as to why it did have a significant effect on the participant's individual performance and not with their collaborative performance. One explanation for this pertains to the obvious differences between the nature of a monologue and a dialogue. In the case of pair work tasks, students had to partly focus on the back-channeling and questions and comments of their interlocutors and provide responses when necessary. This shift of focus reduces the fluency measure. The pauses associated with turn-taking requests, and giving the floor also increases the number of pauses which in turn adversely affects fluency. In contrast, when there is no interlocutor, they will completely focus on their performances and as they are familiar with the task type and its communicative goals, it resembles the first task in such terms, and not familiar with the exact content of the task, they will prioritize meaning over form. The fact that there was no transfer of gains is in congruity with Bygate's (2001) findings that practice did not appear to assist performance on a new task of the same type. In Gass et al. (1999), there was again no transfer of these effects to a new task.

    The third research question of the study focused on whether task repetition and task recycling differ in the effects they have on oral performance of learners. The results gained confirmed that task repetition substantially differs from task recycling in improving the oral performance of learners in terms CAF with significant differences in favor of task repetition. As Bygate (1999, cited in Ellis, 2003) observes, learners are likely to initially focus on message content and subsequently, once message content and the basic language needed to encode it has been established, do they switch attention to the selection and monitoring of appropriate language. Bygate suggests that rehearsal – giving learners the opportunity to repeat a task (Ellis, 2003) – may afford learners the extra processing space they need to integrate the competing demands of fluency, accuracy, and complexity. As Ellis (2003) states, rehearsal can also be seen as a kind of strategic planning also improving the performance of a task, especially in terms of complexity. It apparently encourages learners to provide more detailed content and to draw more fully on their rule-based system.

    The results of such a task-based research also lend support to hierarchal models of language production of the kind proposed by Levelt (1989):

 

1. Conceptualizing the message;
2. Formulating the language representation; and
3. Articulating the message.

 

    This indicates that competing demands of conceptualization on the one hand and of formulation and articulation on the other are often severe, causing learners to prioritize one aspect of production. This can furnish further explanation as to why in the case of task recycling the only viable change was for fluency and not the other two measures, that is accuracy and complexity.      

    Along with the quantitative analysis of the data in the present study, there was an informal investigation of the data gathered which revealed certain interesting notions. There were as many as 28 instances when an error or a mistake spotted in the first performance was not made a second time during task repetition since it simply was avoided because of the carry over to the second performance. More interestingly, there were many instances where one specific word, phrase or even a grammatical structure, e.g. passive voices and participial phrases, was overheard and then borrowed and used. The majority of this borrowing was lexical and even here the task repetition group was superior with 19 instances as opposed to only 9 cases in task recycling where a particular mistake was corrected as a result of being used by the partner or some other student in the classroom while reporting to class during the first task performance. What follows illustrates two very interesting cases of borrowing from the already produced language: One involving a syntactic one and the other a lexical one. One of the students in task the repetition group (Mohamad), who was the most fluent member of the group, with the IELTS speaking band score of 7, seemed to particularly enjoy this opportunity, with six instances, four of which were lexical, to benefit from the various alternatives the task repetition offers to vary his structures in the second performance. The phrase in bracket is the native correct form or use. Triple dots indicate pauses or interruptions by the interlocutor:

(NB. for ethical considerations, pseudonyms are used in the following sample performances.)

 

Lexical

First performance (on crossing culture task)

Aziz: I guess [um] … there should some sort of trade-off between local culture (the local)

Mohamad: How do you mean?

Aziz: I mean, since when we are living in a global village, there should a balance between the local and the international customs.

 

 

Second performance

Aziz: I don't think a big part of (a major part of) traditional culture is gonna be useful for… modern societies.

Mohamad: I agree that many of them aren't useful, but I think there should be a trade-off between modernity and tradition. We can't get rid of it (we can't ignore it).

 

    One interesting point is Mohamad’s exact use of the modal auxiliary as Aziz did. He, in a way, tried to be faithful to the original combination for the fear that he might make a mistake as long as collocation is concerned. More research is needed to determine this conservative use of modal auxiliary is just one single case or general trend as it was also spotted in four other cases with the lexical modifications.

 

Syntactic

    Ahmed was one of the rather less proficient students compared with others as he gained the IELTS speaking band score of 6 prior to the beginning of the study. The following excerpt depicts how he corrects his use of participial forms as he sees it used in the utterance of one of the students in the classroom:

 

First performance (on job characteristics)

Babak: I'm sure a salesman like him meets a lot of famous people [um] like actors.

Ahmed: …yeah, but that mean (means) his be on his feet a lot.

Babak: His being on foot is worth it, because he can make a lot of new friends who are fashionable [laughter]

 

Second performance (on job hunting)

Ahmed: I don't like a job like a walking tour guide because it mean (means) his being on his feet and it is very difficult.

 

    Although short, this excerpt gives a number of insights. One interesting point is how Ahmed managed to pick up a correct participial form from Babak’s utterance, but he was not able to correct his “mean” (means) in spite of the fact that he used it twice. One other issue which is quite ambiguous is the certainty with which we can expect to see such borrowings if for example Babak had not immediately used the structure because Ahmed might have simply missed the correct use of it if it had not been used right after his own production. It deserves more investigation as to why he repetitively had problem with the third person use of verbs which is comparatively a much simpler structure than participial forms that he instantly picked up and correctly used.

 

CONCLUSION AND IMPLICATIONS

Although tasks and TBLT have dominated much of English language teaching and SLA spheres over the last thirty years, there are still certain major controversies as to what exactly a task is and how it can be best implemented in order to promote CAF. In fact, tasks have been more of a research tool rather than a teaching device which in turn justifies the existence of a set of practical gaps in using them, which strongly opposes the rich theoretical background they enjoy. This study was carried out in order to investigate the effects of task repetition and task recycling, as two options the post-task stage of the task cycle offers, on oral performance of learners in terms of fluency, accuracy, and level of complexity on both individual and collaborative levels. A within group as well as between group design was implemented with two experimental groups and no control group. The groups performed under identical conditions in terms of the physical characteristics of the classroom, the instructor (the second researcher), instructions, time allotment and corrective feedback which were none. On each second performance of tasks in both groups toward the end of class, the students were asked to do the task a second time under the pretext that the first performance had not been satisfactory. No elaboration, however, was furnished on their syntactic, lexical or pragmatic errors or mistakes, whatsoever. On no occasion did the participants have any practice or opportunity to prepare themselves for the task since the instruction for the second performance was totally unexpected. The participants' performances were recorded during the four consecutive data collection process and were then transcribed. Concerning the first research question, the results revealed a significant effect task repetition had on performance measures of fluency, accuracy and complexity. With regard to the second research question, the results did not show a meaningful change in oral performance of learners. The only measure on which task recycling did have a viable impact was fluency in individual performances and overall performance, but not the pair work. Finally, as for the third research question, the results revealed a significant difference between task repetition and task recycling on how they influenced oral performance. It was concluded that task repetition outweighed task recycling in improving the CAF measures.    

    The results gained from the present study are of paramount significance and in fact can benefit SLA researchers, course designers, materials developers, test constructors, and language teachers since finding the appropriate measures and methods elaborating on how to hone productive skills and in particular the oral performance of learners has always been a great concern for all teaching methodologies over years. More importantly, one of the major criticisms leveled at TBLT has always been concerned with whether tasks can help develop fluency at the expense of accuracy. The independent variables of the study, task repetition and task recycling are two good options on how learners could consolidate uptakes from doing tasks.

    As for materials developers and course designers, they can clearly benefit from the findings of the present study as they guide what task types can be added to post-task stage of the task cycle to ameliorate learners' performance. Besides, given the novelty of the work done on task recycling effects and its differences with task repetition, they now have a clearer picture of what task recycling and repetition have to offer in practice; course designers and materials writers can incorporate such activities in the syllabus as the need arises. For instance, for lower level learners where there is more focus on their fluency, it is quite logical to include task repetition as well as task recycling, especially individual task recyclings, in the syllabus while with intermediate students or upper intermediate students who have already developed fluency, but are not accurate enough, task repetition can prove to be useful in helping them improve accuracy and complexity.

    Teachers and test constructors can as well enjoy the benefits of the present results. Teachers can use the results in that they know what sorts of options they have at their disposal and what each can afford to better enhance students learning, and, more precisely, enhance learners' oral performance and what it is exactly that improves. This is because it can be quite disappointing for both teachers and learners when teachers’ expectations prior to task implementation are not catered for by what the task and its phases have to offer in reality. Test constructors can gain insight from the results since they can predict the learners' behavior on certain testing situations and can construct tests accordingly whenever a desired performance is in mind.  It is worth mentioning that there is a trade-off between meaning and form and that, given the limited attentional resources, learners usually favor attention to meaning which sacrifices accuracy and complexity. This point can clearly explain the performance differences observed within one learner in different conditions.

    SLA researchers appear to benefit most from the results of the present study as it unclouds certain mysteries and ambiguities about task recycling. As mentioned earlier, there is very little said about task recycling and how it actually benefits learners if utilized. Concerning task repetition, as well, there are certain new gains contrary to the results of previous studies that task repetition does improve learners' oral performance in all its three measures of complexity, accuracy, and fluency.

تعداد مشاهده مقاله: 5,977

تعداد دریافت فایل اصل مقاله: 3,461